WARN: Write to InfluxDB failed (attempt: 1). y [RequestTimedOutError]: Request timed out

[janvda]

When writing at a very high rate measurements to my influx DB (on another device) I am getting the following error. Moreover when having that error node-red seems to hang for a while.

WARN: Write to InfluxDB failed (attempt: 1). y [RequestTimedOutError]: Request timed out
    at ClientRequest.<anonymous> (/data/node_modules/@influxdata/influxdb-client/dist/index.js:16:10715)
    at ClientRequest.emit (events.js:400:28)
    at Socket.emitRequestTimeout (_http_client.js:790:9)
    at Object.onceWrapper (events.js:519:28)
    at Socket.emit (events.js:412:35)
    at Socket._onTimeout (net.js:495:8)
    at listOnTimeout (internal/timers.js:557:17)
    at processTimers (internal/timers.js:500:7)
25 Nov 15:28:44 - [info] [debug:catch all] 
{
  topic: '/miband-api/activity',
  payload: [
    [
      { category: 112, intensity: 0, steps: 0, time: 1637474340000 },
      { person: 'jan' }
    ]
  ],
  qos: 0,
  retain: false,
  _msgid: '6e57bd75ebca7689',
  ttl: 0,
  _queuetimestamp: 1637850474084,
  _queueCount: 4215,
  last_activity_time_ms: 1637474280000,
  expected_time_ms: 1637474340000,
  influx_error: {
    errorMessage: y [RequestTimedOutError]: Request timed out
        at ClientRequest.<anonymous> (/data/node_modules/@influxdata/influxdb-client/dist/index.js:16:10715)
        at ClientRequest.emit (events.js:400:28)
        at Socket.emitRequestTimeout (_http_client.js:790:9)
        at Object.onceWrapper (events.js:519:28)
        at Socket.emit (events.js:412:35)
        at Socket._onTimeout (net.js:495:8)
        at listOnTimeout (internal/timers.js:557:17)
        at processTimers (internal/timers.js:500:7)
  },
  error: {
    message: 'RequestTimedOutError: Request timed out',
    source: {
      id: 'b1dbea7902f83c6d',
      type: 'influxdb out',
      name: 'my influx db - activity measurement',
      count: 1
    },
    stack: 'RequestTimedOutError: Request timed out\n' +
      '    at ClientRequest.<anonymous> (/data/node_modules/@influxdata/influxdb-client/dist/index.js:16:10715)\n' +
      '    at ClientRequest.emit (events.js:400:28)\n' +
      '    at Socket.emitRequestTimeout (_http_client.js:790:9)\n' +
      '    at Object.onceWrapper (events.js:519:28)\n' +
      '    at Socket.emit (events.js:412:35)\n' +
      '    at Socket._onTimeout (net.js:495:8)\n' +
      '    at listOnTimeout (internal/timers.js:557:17)\n' +
      '    at processTimers (internal/timers.js:500:7)'
  }
}

[janvda]

I understand that this error might be normal when overloading the influx db with write requests but how can I best throttle this ?

One simple solution would be that I write the next measurement once the previous measurement has been sent to the influx DB but the influxdb out node I am using has no output, so that option is not possible.

Maybe instead of sending a single measurement, I should bundle them and write them in one call to the influx db.

[mblackstock]

One way might be to write the next measurement once the node indicates it's done. You could try wiring up a 'complete' node to do this. Once the node indicates it's done, you could trigger the next write. Hopefully this gives you enough to try that out. Here are some references.

• https://nodered.org/blog/2019/09/20/node-done
• https://techexplorations.com/guides/esp32/node-red-esp32-project/node-red-complete/

Another thought may be to use the delay node:

https://cookbook.nodered.org/basic/rate-limit-messages

Hope that helps.